Related Experiment Video
Updated: Jul 16, 2026
Detection of a Circulating MicroRNA Custom Panel in Patients with Metastatic Colorectal Cancer
Published on: March 14, 2019
Plasma Sphingomyelin as a Post-Treatment Monitoring Biomarker for Pathological Response in Locally Advanced Rectal
Pedro Brandão1,2,3,4, Lúcia Lacerda2,3,4,5, Marisa D Santos1,2,3,4
1Colorectal Unit, Department of Surgery, Hospital de Santo António, Unidade Local de Saúde de Santo António (ULSSA), Largo Professor Abel Salazar, 4099-001 Porto, Portugal.
Abstract:
Background: Organ-preservation strategies in locally advanced rectal cancer (LARC) increasingly require accurate assessment of response to neoadjuvant therapy, yet magnetic resonance imaging poorly distinguishes complete from near-complete responses. We investigated plasma sphingolipid dynamics across multiple treatment timepoints as monitoring biomarkers for pathological response. Methods: This single-centre cohort study included 86 patients with histologically confirmed LARC; 58 received neoadjuvant treatment followed by surgical resection with an evaluable Mandard tumour regression grade (TRG) and formed the response-evaluable cohort (19 good responders [TRG 1-2], 39 poor [TRG 3-5]). Plasma sphingomyelin (SM), sphingosine-1-phosphate and glucosylceramide were quantified by UPLC-MS/MS at baseline (M0), post-chemoradiotherapy (M1) and post-surgery (M2). Discrimination was assessed by receiver operating characteristic analysis and odds ratios, longitudinal trajectories by a linear mixed-effects model, and SM was compared head-to-head with serum carcinoembryonic antigen (CEA). Results: SM did not discriminate response at baseline (M0 AUC = 0.53) but did after treatment (M1 AUC = 0.750, p = 0.024; M2 AUC = 0.786, p = 0.010; pooled AUC = 0.767; odds ratio 10.67 at a 164.0 mg/L cutoff; negative predictive value 84%). SM correlated inversely with TRG grade (Spearman ρ = -0.43, p = 0.016) and outperformed serum CEA (AUC 0.74 vs. 0.60). Trajectories diverged over time (mixed-model time × group interaction p = 0.023) with no baseline difference (p = 0.96): good responders increased SM by +22.5% from M0 to M2, whereas poor responders declined by -8.4%. Conclusions: Plasma SM is not a pre-treatment predictor but emerges as a post-treatment monitoring biomarker for pathological response in LARC, supporting its integration into multimodal response assessment for organ-preservation decision-making. External validation in independent cohorts is warranted.
